Network performance tests are useful for evaluating the performance of various nodes in a network. Such performance tests typically have involved sending test packets through the network for evaluating network performance. Existing performance tests, however, have determined the status of a given node in the network by inferring from errors in the network performance tests, or from measures of the central processing unit (CPU), memory or other vital signs of the node. Such existing approaches to network performance measurement lead to confusion since the errors in the network performance tests may have been caused by many other network problems, and not necessarily by the status of a given network node.